Logo Utrecht University

S4 conference

Pre/post-conference workshops

Pre-conference workshops (March 5, 2018)

Gentle Introduction to Bayesian Analysis with Small Samples

Instructor: Rens van de Schoot

Bayesian methods are becoming increasingly popular as a solution for small sample problems in social sciences. Despite their popularity, these methods have yet to become a standard part of the statistics curricula in graduate programs.

This workshop is designed for applied researchers who are new to Bayesian methods and would like to learn the theory behind Bayesian statistics, the differences between Bayesian and frequentist statistics, and how to apply Bayesian methods to answer their research questions. The focus of the workshop will be on how Bayesian methods can be used in small sample research.

 

Mediation Analysis with Small Samples

Instructor: Milica Miočević

Mediation analysis is used to evaluate the mechanism through which the independent variable(s) affect the dependent variable(s). Bayesian methods with accurate prior distributions were found to increase power to detect the mediated effect in small samples.

The workshop starts with a tutorial on how to use Bayesian methods in linear models, and proceeds to cover two ways to do Bayesian mediation analysis. This workshop is designed for researchers who are familiar with the theory behind Bayesian statistics and linear regression analysis. Participants new to Bayesian methods are encouraged to first attend the workshop “Gentle Introduction to Bayesian Analysis with Small Samples”.

 

Post-conference workshops (March 8, 2018)

Multilevel Modeling with Small Samples

Instructor: Dan McNeish

Multilevel models (MLMs) are used to analyze data that have a nested (hierarchical) structure, e.g., students are nested within classrooms. In small samples MLMs encounter convergence issues and yield parameter estimates with poor statistical properties.

This workshop focuses on available methods for avoiding such issues. The target audience for this workshop are researchers in social sciences who work with nested data and small samples. Participants new to Bayesian methods are encouraged to first attend the workshop “Gentle Introduction to Bayesian Analysis with Small Samples”.

 

Latent Growth Curve Modeling with Small Samples

Instructor: Sarah Depaoli

Latent growth curve models (LGMs) are used to model the changes in a construct over time. In small samples LGMs encounter convergence issues and yield parameter estimates with poor statistical properties.

Bayesian methods with informative prior distributions offer a solution to these issues. The target audience for this workshop are social sciences researchers who work with longitudinal data and small samples. Participants new to Bayesian methods are encouraged to first attend the workshop “Gentle Introduction to Bayesian Analysis with Small Samples”.